1. Field of Invention
The present invention relates to an improved method for preparing oxycodone. More particularly, the present invention sets forth a method for preparing oxycodone in high yields that does not require the employment, or synthesis, of thebaine in the reaction scheme.
2. Background of the Related Art
The analgesic activity of Papaver somniferum has been known since antiquity. It has long been understood that the milky juice derived from the unripe seed capsules of this poppy plant possesses potent pharmacological properties. The dried and powdered form of the juice is referred to as opium. Opium comprises about 10% of the juice obtained from the unripe seed capsules of Papaver somniferum. 
Early in the nineteenth century it was recognized that opium contains numerous alkaloid compounds. The first of these alkaloids to be isolated was morphine, described by Serturner in 1805. The isolation of other alkaloids, including codeine (Robiquet 1832), papaverine (Merck 1848), thebaine, oripavine and noscapine followed in short order. By the middle of the nineteenth century, the use of pure alkaloids rather than crude opium preparations was established medical practice. It is now known that opium contains more than twenty distinct alkaloids.
In general, the opium alkaloids can be divided into five distinct chemical classes: phenanthrene, benzylisoquinoline, tetrahydroisoquinoline, cryptopine and miscellaneous (Remington""s Pharmaceutical Sciences 433, 1975). Therapeutically useful drugs are primarily isolated from the phenanthrene and benzylisoquinoline classes. The principal phenanthrenes are morphine (≈10% of opium), codeine (≈0.5% of opium) and thebaine (≈0.2% of opium). The principal benzylisoquinolines are papaverine (≈1.0% of opium) and noscapine (≈6.0% of opium).
Morphine itself comprises a 5-ring structure, incorporating a partially hydrogenated phenanthrene ring system. Each ring of morphine is designated as set forth below: 
Morphine includes what is referred to in the art as a morphinan ring structure, comprising rings A, B, C and E, as set forth below: 
The substituent numbering of morphine derivatives follows two common conventions as shown: 
It is the second (Chemical Abstracts) numbering system that shall be made reference to hereinafter.
The first total synthesis of morphine was published in 1952 (Gates, et al., 74 J. Amer. Chem. Soc., 1109, 1952). Because the laboratory synthesis of morphine is difficult, however, the drug is still obtained from opium or extracted from poppy straw (Goodman and Gilman""s The Pharmacological Basis of Therapeutics, 489, 1990). Semi-synthetic derivatives of the naturally occurring opium alkaloids are widely employed in medicine today. Among the important properties of opioids that may be altered by structural modification are the affinity of the compound for various species of opioid receptors, resistance to metabolic breakdown, lipid solubility and agonist versus antagonist activity.
Codeine, hydrocodone, hydromorphone, oxycodone, and oxymorphone which are found in present day analgesic prescription drugs, are all congeners of morphine. Other structural analogs of morphine used medically in the United States include: levorphanol, nalmefene, naloxone, naltrexone, buprenorphine, butorphanol, and nalbuphine. Some morphine analogs, such as levorphanol, may be produced totally synthetically around a non-opiate morphinan nucleus which is synthesizable from coal tar derivatives (Remington""s Pharmaceutical Sciences 1039, 1975).
Among the many morphine structural analogs used in medicine today, widespread use is made of both codeine and oxycodone.
Codeine is 3-methylated morphine. Codeine has less than one-seventh the analgesic potency of morphine (Foye, Medicinal Chemistry, 254 (1975)). However, as codeine has a far better oral bioavailability than morphine (the 3-methoxy group is believed to protect it from rapid first-pass biotransformationxe2x80x94the action of morphine orally is terminated largely by glucuronide conjugation at the 3-hydroxyl group), codeine is only less than four times as potent, on a weight basis, than morphine when both compounds are administered orally (Drug Facts and Comparisons 1246, 1996). While some codeine is obtained from opium directly, the quantity obtainable from such extraction is not sufficient to meet the extensive use of the alkaloid. The need for codeine is fulfilled by partial synthesis of the compound from morphine (Remington""s Pharmaceutical Sciences 1038, 1975).
Oxycodone is a white, odorless crystalline powder of semi-synthetic origin with multiple actions qualitatively similar to those of morphine. 
The principal actions of therapeutic value are analgesia and sedation. It is similar to codeine and methadone in that it retains at least one half of its analgesic activity when administered orally. It is a pure agonist opioid, which produces not only analgesia, but other therapeutic effects including anxiolysis, depression of the cough reflex, euphoria and feelings of relaxation. On a weight basis, oxycodone is approximately twice as potent orally as morphine (Drug Facts and Comparisons 1246, 1996). Oxycodone is typically indicated for the relief of moderate to moderately severe pain (Drug Facts and Comparisons 1259, 1996).
Thebaine, which also contains a morphinan-ring structure, differs from codeine in replacing the hydroxyl group of the morphinan C-ring with a methoxy group and the xe2x80x9cCxe2x80x9d ring has two double bondsxe2x80x94xcex946,7, xcex948,14 .(i.e., thebaine differs from morphine in that both hydroxyl groups are methylated and the xe2x80x9cCxe2x80x9d ring has two double bondsxe2x80x94xcex946,7, xcex948,14). 
The compound demonstrates the effect that minor modifications in structure of morphinan compounds may have in pharmacological effects, as thebaine lacks any substantial analgesic activity (Foye, Medicinal Chemistry, 256 (1975)).
While lacking medicinal usefulness in itself, thebaine is singularly important as a key intermediate in the synthesis of many useful opiate-derivatives (See, Barber et al., 18 J. Med. Chem. 1074-107, 1975), including oxycodone (Freund et al., 94 J. Prak. Chemie 135-178, 153, 1916; See Physician""s Desk Reference, 2569, 54th Ed. 1999), naloxone, naltrexone and nalbuphine (See, U.S. Pat. No. 4,795,813 at Col. 1, lines 16-21). Thebaine is the only known xcex946,8-diene compound among the naturally-ocurring morphine alkaloids (Seiki, 18 Chem. Pharm. Bull. 671-675, 1970).
Oxycodone may be prepared from thebaine by: dissolution of the thebaine in aqueous formic acid, oxidation treatment with 30% hydrogen peroxide (Seki, 18 Chem. Pharm. Bull. 671-676, 1970), neutralization with aqueous ammonia to yield 14-hydroxycodeinone and hydrogenation of the 14-hydroxycodeinone in acetic acid with the aid of a palladium-charcoal catalyst (Remington""s Pharmaceutical Sciences 1041, 1975). Oxidation of thebaine may alternatively be performed using potassium dichromate in acetic acid (Freund et al., 94 J. Prakt. Chem. 135, 1916) or performic acid (Iljima et al., 60 Helv. Chim. Acta 2135-2137, 1977). Improved yield, however, has been reported to be obtained by oxidizing with m-chloroperbenzoic acid in an acetic acid-trifluoroacetic acid mixture (Hauser et al., 17 J. Med. Chem. 1117, 1974; See also, U.S. Pat. No. 4,795,813 to Schwartz, Col. 1, Lines 22-26). Yield may also be improved by hydrogenation of 14-hydroxycodeinone under a pressure of about 30 psi (Kraxcex2nig et al. 329 Arch. Pharm. Pharm. Med. Chem. 325-326, 1996).
Although particularly useful in the synthesis of numerous pharmaceutical preparations, thebaine is among the least abundant phenanthrene alkaloids in Papaver somniferum. Due to its scarcity, a number of investigators have proposed methods of obtaining this unique alkaloid using other more abundant opioid compounds as starting materials.
Seki (18 Chem. Pharm. Bull. 671-676, 1970) discloses a method for preparing xcex946,8-diene compounds, such as thebaine, from xcex1,xcex2-unsaturated ketones such as codeinone, which may be obtained from the natural alkaloid codeine. Codeinone was added to a mixture of p-toluenesulfonic acid (dehydrated prior to reaction), absolute methanol and dried benzene, the solution refluxed for 3 hours under azeotropic removal of water, and the reaction mixture purified by washing with diluted sodium hydroxide, to obtain thebaine. A reported maximum yield of 26.8% was reported when using 1.1-0.15 molar equivalents of p-toluenesulfonic acid to codeinone. Eppenberger et al. (51 Helv. Chim. Acta 381, 1968) report a four step method for converting dihydrocodeinone to thebaine which results in a similar yield of 27%. Schwartz et al. (97 J. Am. Chem. Soc. 1239, 1975) demonstrate the total synthesis of thebaine in which the key step is the oxidative coupling of a reticuline derivative to a salutaridine derivative. The overall yield of dl-thebaine, however, was only in the 1-2% range based on isovanillin. Reaction of salutaridinol with an organic or inorganic acid halide or acid anhydride, followed by treatment with a strong base, is taught as a method of thebaine production in U.S. Pat. No. 3,894,026 to Sohar et al. A yield as high as 50.3% was reported (See, Col. 4, Line 29). Barber et al. (18 J. Med. Chem. 1074-1077, 1975) report synthesizing thebaine (as well as oripavine) from codeine and morphine. Barber et al. teach methylation of the potassium salt of codeine to give codeine methyl ether followed by oxidation with xcex3-MnO2 (See also, U.S. Pat. No. 4,045,440 to Rapoport et al., 1977). These authors claim a 67% yield of oxycodone from codeine. European Patent Application No. EP 0 889 045 A1 likewise teaches a process for the production of thebaine from the more readily available morphinans codeine and morphine. Such method provides for converting the starting material to an alkali metal or quaternary ammonium cation and reacting the same with a compound of the formula RX wherein R is an alkyl or acyl group and X is a leaving group.
While all of the above methods have been devised to increase the supply of thebaine by synthetic and semi-synthetic means, the fact remains that thebaine remains relatively costly as opposed to morphine and codeine.
The use of thebaine as a starting material to form other therapeutically useful opioids also suffers from a disadvantage unassociated with its relative scarcityxe2x80x94thebaine is a known convulsant, capable of causing (even in low doses) strychnine-like convulsions (Foye, Principles of Medicinal Chemistry 255, 1975; The Merck Index, 9203 (11th Edition), 1989). Employment of thebaine in any synthesis scheme, therefore, entails significant risks and requires the taking of a number of precautions. Considering the relatively high cost of, and the toxicity potential of, thebaine, it would be preferred if alternative synthesis methods were developed to manufacture the many opioid congeners currently synthesized from thebaine from cheaper and less toxic materials.
U.S. Pat. No. 2,654,756 discloses a method for converting codeine into codeinone, dihydrocodeinone and dihydromorphine rather than synthesizing such compounds from thebaine. Conversion is effectuated by way of oxidation using certain ketones in the presence of aluminum alkoxides. Likewise, methods for producing 14-hydroxymorphinans, such as naloxone, naltrexone and nalbuphine (opioid antagonists) from codeine, without a thebaine intermediate, have also been disclosed (See, U.S. Pat. No. 4,472,253 to Schwarz and Schwartz and Wallace, 24 J. Med. Chem. 1525-1528, 1981). To date, however, no economical method has been proposed for manufacturing oxycodone from a readily available starting material that has a toxicity and cost profile which is significantly improved over that possessed by thebaine.
The present invention provides an improved, high-yield, method for preparing oxycodone that does not require employment of a thebaine intermediate in the reaction scheme. The disclosed method makes use of compound having a morphinan-like ring structure, such as codeine or morphine, as a starting material for the synthesis of oxycodone. The method employs the steps of: converting the starting material to a compound with a morphinone ring structure, preparing a dienolsilyl ether at the C-ring of the morphinan-like ring structure by reacting an organosilyl compound with the starting material, oxidizing the silyl ether, and hydrogenating the unsaturation in the C-ring. Formation of the dienolsilyl ether is promoted by efficient dienolization of the C-ring, which is provided by reacting the xcex1,xcex2-unsaturated ketone and organochlorosilane reactants in the presence of a strong amine base, such as DBU (1,8-Diazabicyclo[5.4.0.]undec-7-ene) or DBN (1,5-Diazabicyclo[4.3.0]non-5-ene). Preferably the organosilyl reactant, for example a triorganosilyl chloride, is stericly hindered on the silicone atom.
An aspect of the present invention comprises a method for producing oxycodone from codeine employing two oxidation steps, one involved in the oxidation of a hydroxyl group to a ketone, and the other involving oxidative hydroxylation of a dienolsilyl ether. In particular, oxycodone free base has been produced in commercially reasonable yields by forming a dienolsilyl ether derivative of codienone in the presence of a strong amine base (preferably a diazabicyclo-base), oxidizing the silyl ether to form 14-hydroxycodeinone, and hydrogenation of the morphinan C-ring unsaturation to form oxycodone.
In one embodiment of the present invention, there is provided an improved method for synthesizing oxycodone from codeine free base. In this embodiment, codeine free base is converted to codeinone by oxidation, for example, by using a standard oxidant such as MnO2, Na2WO4/H2O2, Pd(OAc)2/O2, and/or a standard oxidation procedure, e.g., Swern/Moffat-type oxidation (DMSO-based oxidation), Oppenauer-type oxidation (employing aluminum alkoxides and cyclohexanone or other ketones). Preferred oxidants include BaMnO4 and Oppenauer oxidation. Codeinone is then reacted with an organosilyl compound having an effective leaving group, such as a halogen. The resulting dienolsilyl ether derivative is then oxidized with an oxidizing agent to afford 14-hydroxycodeinone. It has been found that the dienolsilyl ether of the morphinone C-ring may efficiently be converted to 14-hydroxycodeinone using peracetic acid solution. Hydrogenation of the unsaturation in the C-ring is subsequently performed and may be accomplished by way of, for example, pressurized catalytic hydrogenation or catalytic transfer hydrogenation in acetic acid. Oxycodone produced by such method has been found to be obtainable in yields approximating 80%.
One of the novelties of this invention is the discovery that commercially-practicable yields of therapeutically employed opioid alkaloids having a morphinan ring structure can be obtained without recourse to a thebaine intermediate by reacting a compound with a morphinone ring structure with an organosilyl reactant in the presence of a strong amine base, preferably a diazabicyclo-base such as DBU (1,8-Diazabicyclo[5.4.0.]undec-7-ene) or DBN (1,5-Diazabicyclo[4.3.0]non-5-ene) (to improve enolization and the promotion of a dienolsilyl ether derivative), followed by oxidation of the dienolsilyl ether moiety.
After considerable experimentation with numerous reaction schemes designed to form oxycodone from codeine and morphine (two relatively inexpensive opioid alkaloids), the present inventors have discovered a unique reaction scheme for manufacturing oxycodone that provides for industrially-acceptable yields. The present invention overcomes many of the prior art problems associated with the production of oxycodone and provides for a synthetic scheme for oxycodone production which does not employ the relatively costly, scarce and toxic alkaloidxe2x80x94thebaine.
The present inventors have discovered that enolization of the C-ring of a morphinone compound having an xcex1,xcex2-unsaturated ketone structure, is significantly enhanced by exposure to a strong amine base, such as DBU or DBN and similar diazabicyclo-bases. Formation of the dienolsilyl ether (by reaction with the ketone of such ring with a organosilyl compound having an effective leaving group) was greatly improved by effectuating the reaction in the presence of the strong amine base. The present inventors have further discovered that the dienolsilyl ether of codeinone (the silyl ether formed at position 6 (chemical abstract substituent-numbering designation)) may be used to directly form 14-hydroxycodeinone by oxidation of the silyl ether. In a preferred embodiment, oxidation is performed at room temperature for about 3 hours. The dienolsilyl ether of codeinone may be dissolved in toluene or other similar solvent. Oxidation may be efficiently performed with relatively high yield using peracetic acid or other peracids. Hydrogenation of 14-hydroxycodeinone produces oxycodone. A preferred hydrogenation reaction employs hydrogen gas or NaH2PO2 along with a palladium-carbon catalyst, with the 14-hydroxycodeinone being dissolved in a weakly acidic solution such as aqueous acetic acid.
A codeinone dienol silyl ether, such as the intermediate compound formed in the conversion of codeine to oxycodone according to certain embodiments of the present invention, is disclosed in pending European patent application No. EP 0 889 045 A1 to Jen-Sen Dung. The reference, however, is instructive as to the unobviousness of the present invention.
Recognizing the expense and relative scarcity of thebaine, EP 0 889 045 A1 teaches (as noted above) a process for the production of thebaine and analogues thereof. While disclosing codeinone tert-butyldimethylsilyl dienol ether (Example 6), the patent teaches the production of oxycodone only from thebaine which is synthesized by the procedures described (See, e.g., Abstract of the Disclosure, Col. 1, Lines 25-52, Col. 5, Lines 24-29, Col. 9, Example 8). No recognition is made of the fact that the tert-butyldimethylsilyl dienol ether could be utilized, without synthesis of a thebaine intermediate, to produce oxycodone. Further the reference fails to teach a method for producing organosilyl dienol ethers of the morphinone ring in commercially practicable yields. The reference notes that the codeine tert-butyldimethylsilyl dienol ether produced by the methods described comprised only 23% of the solid mass recovered (thus comprising a relatively minor component of the solid mass). EP 0 889 045 A1 does not disclose or imply that yield could be significantly enhanced by the presence of strong amine base (rather than tetrahydrofuran as taught by the reference) in the reaction mix when the ether is being formed.
The presently disclosed invention provides commercially practicable yields, yields typically in excess of 50%, and more typically in excess of 80%, of oxycodone from codeinone (a compound that is easily obtained from codeine by oxidation). Codienone is easily synthesized from codeine, an alkaloid that can be obtained naturally, or semi-synthetically, as from morphine. It has been discovered that by reacting an organosilyl compound in the presence of a strong amine base that a high degree of conversion to the organosilyl dienol ether conjugate of codeinone may be achieved. The strong amine base is believed to strong favor enolization of codeinone, a compound having an xcex1,xcex2-unsaturation in the xe2x80x9cCxe2x80x9d ring of the morphinone ring structure, while the organosilyl moiety captures the enol form. The organosilyl ether form of codeinone is also promoted by employing an organosilyl compound having an effective leaving group, such as a halogen, and in employing a stericly bulky silicone moiety. While the resulting dienolsilyl ether form of codeinone may be oxidized to 14-hydroxycodeinone using a number of standard oxidizing agents, it has been found that oxidation with peracetic acid is extremely efficient, producing about an 80% yield. The 14-hydroxycodeinone is then hydrogenated, as by catalytic hydrogenation, so as to hydrogenate the xcex1,xcex2-unsaturation of the C-ring. A catalytic transfer hydrogenation method in aqueous acetic acid was found to produce about the same yield, and similar impurity patterns, as the method reported by R. Kraxcex2nig, et al.
In an aspect of the invention, there is disclosed a method of producing oxycodone from codeinone which comprises the steps of: (a) producing a dienol organosilyl ether at position 6 of the C-ring of codeinone thereby forming a dienol organosilyl ether congener of codeinone; (b) oxidizing the dienol organosilyl ether to form 14-hydroxycodeinone; (c) hydrogenating the unsaturation in the C-ring of 14-hydroxycodeinone to produce oxycodone.
The dienol organosilyl ether congener of codeinone is preferably formed by reacting an organosilyl compound with codeinone, such organosilyl compound having the formula:
R33SiX 
wherein R3 is alkyl or aryl and the three R3 groups may be the same or different and X is a leaving group, such as imidazole, mesylate, tosylate or a halogen. Preferably, the organosilyl compound is reacted with codeinone in the presence of a strong amine base, such as diazobicyclo-base, for example, DBU (1,8-Diazabicyclo[5.4.0.]undec-7-ene) or DBN (1,5-Diazabicyclo[4.3.0]non-5-ene). Oxidation of the dienol organosilyl ether may be performed by treating the dienol organosilyl ether congener of codeinone with peracetic acid, preferably in the presence of an organic solvent such as toluene.
In another aspect of the present invention there is disclosed a method for oxidizing a dienol silyl ether selected from the group having the formula: 
wherein R1 is selected from the group of alkyl or acyl and R2 is selected from the group of lower alkyl, allyl, or lower alkyl substituted by cycloalkyl, and R3 is an alkyl or aryl group and the three R3 groups may be the same or different, which comprises the steps of: (a) reacting the dienol silyl ether compound with peracetic acid and (b) thereafter a work up procedure to isolate the product as a free base.
In yet another aspect of the present invention there is disclosed a method for forming a dienol silyl ether selected from the group having the formula: 
wherein R1 is selected from the group of alkyl or acyl and R2 is selected from the group of lower alkyl, allyl, or lower alkyl substituted by cycloalkyl, and R3 is an alkyl or aryl group and the three R3 groups may be the same or different, which comprises the steps of:
reacting an morphinan-6-one selected from the group having the formula: 
with an organosilyl compound having the formula
R33SiX 
wherein R3 is an alkyl or aryl group and the three R3 groups may be the same or different group and X is a leaving group, such as imidazole, mesyate, tosylate or a halogen, in the presence of a strong amine base. The strong amine base may be a diazabicyclo-base, and may more specifically be selected from the group consisting of: DBU (1,8-Diazabicyclo[5.4.0.]undec-7-ene) and DBN (1,5-Diazabicyclo[4.3.0]non-5-ene). X is preferably chloride.
And yet another aspect of the present invention entails a method of producing oxycodone from codeine which comprises the steps of: (a) oxidizing codeine to codeinone; (b) producing a dienol organosilyl ether at position 6 of the C-ring of codienone thereby forming a dienol organosilyl ether congener of codeinone; (c) oxidizing the dienol organosilyl ether to form 14-hydroxycodeinone; (d) hydrogenating the unsaturation in the C-ring of 14-hydroxycodeinone to produce oxycodone. The dienol organosilyl ether congener of codeinone of this embodiment may be formed by reacting an organosilyl halide with codeinone, preferably an organosilyl halide having the formula: R33SiCl, wherein R3 is as defined hereinabove. Preferably the organosilyl chloride is reacted with codeinone in the presence of a strong amine base. The strong amine base may be a diazabicyclo-base and may be selected from the group consisting of: DBU (1,8-Diazabicyclo[5.4.0.]undec-7-ene) or DBN (1,5-Diazabicyclo[4.3.0]non-5-ene). The oxidation of the dienol organosilyl ether may be performed by treating the dienol organosilyl ether congener of codeinone with peracetic acid (which reaction may be carried out in the presence of an organic solvent such as toluene).
A preferred method of the present invention for forming oxycodone from codeine fundamentally involves four (4) synthetic steps: (1) oxidation of codeine to codeinone; (2) formation of an organosilyl ether congener of codeinone; (3) oxidation of the silyl ether to 14-hydroxycodeinone; and (4) hydrogenation of the partially unsaturated non-aromatic C-ring to produce oxycodone, such as described in more detail below and as shown in the following diagrammatic form: 
The oxidation of codeine to codeinone may be performed by numerous methods known to those of ordinary skill in the art including: CrO3/TBHP oxidation, dichromate oxidation, Na2WO4/peroxide oxidation, BaFeO4 oxidation, hydrous ZrO2/ketone oxidation, oxidation using CrO2, Highet oxidation (Highet et al., 77 J. Am. Chem. Soc. 4399, 1955) using manganese dioxide, Oppenauer oxidation using aluminum isopropoxide and cyclohexanone (See, U.S. Pat. No. 2,654,756 to Homeyer et al.), sodium tungstate activated peroxide oxidation (Sato et al., 119 J. Am. Chem. Soc. 12386, 1997), Swern/Moffatt type (DMSO-based) oxidation, palladium acetate catalyzed aerobic oxidation and barium manganate oxidation (See, Nishimura et al., 39 Tet. Let. 6011).
As would be understood by one of ordinary skill in the art, with respect to any oxidation procedure, adjustment of reaction conditions, such as the concentration of the reactants, the acidity of the reaction mixture, the presence or absence of solvating agents, and the like, .may impact upon the yield of oxidized product. For example, with respect to barium manganate oxidation it may be preferred to keep the reaction mixture at about 0xc2x0 C. and to control the polarity of the solvent to improve yield. With respect to Oppenauer oxidation, the addition of toluene to the reaction scheme may improve yield, as well may azeotropic removal of water from the codeine/toluene solution prior to the addition of a catalytic amount of aluminum isopropoxide, and the collection of distillate during and after the addition of the aluminum isopropoxide. Selection may also be made between numerous potential reactants such as in Swern/Moffatt type oxidation numerous activators in DMSO may be employed including: oxalyl chloride, TsCl, P2O5, TFAA, Ac2O, PySO3, Ts2O, SOCl2, DCC, DIPC, cyanuric chloride, ClSO2NCO, (MeSO2)2O, Cl2, hot air, and the like.
Codeinone may be modified to form a silyl ether at position 6 (Chemical Abstracts designation) by reaction with an organosilyl compound R33SiX. Preferred organosilyl compounds were found to be stericly-hindered at the silicon atom and to have chlorine as the leaving group. The enolized codeinone was efficiently trapped with a trialkychlorosilane, such as tert-butyldimethylchlorosilane or triethylchlorosilane. The trimethylsilyl ether, however, was found to be rapidly hydrolyzed. Enolization of codeinone, and the formation of the dienolsilyl ether, was found to be promoted by the presence of strong amine base, such as DBU or DBN. Other bases such as LDA, DABCO, DIPEA, TEA, imidazole, N-methylmorpholine, HMDS-Li salt, hexamethyldisilazane, and aluminum isopropoxide did not yield a desirable amount of dienolsilyl ether.
Oxidation of the dienolsilyl ether of codeinone to 14-hydroxycodeinone may be performed using the many oxidizing agents and methods known in the art. For example, the dienolsilyl ether may be oxidized in a hydrogen peroxide-free performic acid mixture according to the method published by Swern (D. Swern, Organic Reactions VII, 378, 1953), by way of MnO2 or performic acid. A preferred oxidation procedure, however, was found to employ peracetic acid prepared from acetic anhydride, hydrogen peroxide and a catalytic amount of sulfuric acid. Aging of the peracetic acid solution and treatment with acetic anhydride was found to improve optimum oxidation hydroxylation) presumably by removal of any free hydrogen peroxide. Anhydrous peracetic acid up to 25 days old was found to be most effective. The yield of 14-hydroxycodeinone was found to be also effected by the molar ratio and percentage of oxidant in the mixture. Optimal oxidation conditions may vary with different organosilyl ethers. For example, the presence of trifluoroacetic acid (TFA) was found to improve the oxidation of the triethylsilyldienolate of codeinone. The isolation of 14-hydroxycodeinone may entail de-activation of the spent peracetic acid by treating with either sodium hydrogen sulfite or sodium thiosulfate aqueous solution, and removal of acetic acid solvent (in vacuo), and organic neutral by-products like disiloxane or silanol and N-oxide, by acid/base work up procedures. Oxidation of either t-butyldimethylsilyldienolate or triethylsilyldienolate of codeinone may afford a similar yield of 14-hydroxycodeinone (or its acid salt). Oxidation of the triethylsilyl dienol ether and t-butyldimethylsilyl ether of codeine with peracetic acid was found to produce yields of 14-hydroxycodeinone in excess of about 80%.
14-Hydroxycodeinone was converted to oxycodone by hydrogenation of the xcex1,xcex2-unsaturation in the C-ring. Hydrogenation may be performed by using any of the methods known for hydrogenation of 14-hydroxycodeinone to oxycodone. For example, diphenylsilane and Pd(Ph3P)/ZnCl2 may be used to reduce 14-hydroxycodeinone, as may sodium hypophosphite in conjunction with a Pd/C catalyst in aqueous acetic acid, and Pd/C catalytic transfer hydrogenation.